Optical Character Recognition Problem Research Articles

A custom-built deep learning approach for text extraction from identity card images

<span>Information found on an identity card is needed for different essential tasks and manually extracting this information is time consuming, resource exhaustive and may be prone to human error. In this study, an optical character recognition (OCR) approach using deep learning techniques is proposed to automatically extract text related information from the image of an identity card in view of developing an automated client onboarding system. The OCR problem is divided into two main parts. Firstly, a custom-built image segmentation model, based on the U-net architecture, is used to <span>detect the location of the text to be extracted. Secondly, using the location of the identified text fields, a (CRNN) based on long short-term memory (LSTM) cells is trained to recognise the characters and build words. Experimental results, based on the national identity card of the Republic of Mauritius, demonstrate that our approach achieves higher accuracy compared to other studies. Our text detection module has an intersection over union (IOU) measure of 0.70 with a pixel accuracy of 98% for text detection and the text recognition module achieved a mean word recognition accuracy of around 97% on main fields of the identity card.</span></span>

Approaches to the recognition of handwritten letters of the Azerbaijani language and their analysis

Optical character recognition (OCR) computer systems are widely used for recognition of car license plates, for enabling visually impaired people to work with written information, etc. One of the current areas of OCR problems is the recognition of handwritten characters. This study proposes approaches to the construction of combined models using simple and ensemble learning for recognition of handwritten letters of the Azerbaijani language. The models built using the proposed approaches utilized several classes of features. Based on the results of computer experiments, a comparative analysis of the recognition accuracy of the constructed models and features using the proposed recognition methods is conducted.

Research, Direct Construction form Optical Characteristics – OCR using Department Technology

Optical character recognition – using a deep learning technique that describes an overview of optical character recognition, the basic steps in the optical character recognition problem. At the same time, the thesis introduces in detail the Long short term memory (LSTM) deep learning method and its application in optical character recognition. Through testing on a dataset of 10,000 license plate photos, the thesis showed that the application of the LSTM method is quite effective in optical character recognition.

Towards a unified framework for identity documents analysis and recognition

Identity documents recognition is far beyond classical optical character recognition problems. Automated ID document recognition systems are tasked not only with the extraction of editable and transferable data but with performing identity validation and preventing fraud, with an increasingly high cost of error. A significant amount of research is directed to the creation of ID analysis systems with a specific focus for a subset of document types, or a particular mode of image acquisition, however, one of the challenges of the modern world is an increasing demand for identity document recognition from a wide variety of image sources, such as scans, photos, or video frames, as well as in a variety of virtually uncontrolled capturing conditions. In this paper, we describe the scope and context of identity document analysis and recognition problem and its challenges; analyze the existing works on implementing ID document recognition systems; and set a task to construct a unified framework for identity document recognition, which would be applicable for different types of image sources and capturing conditions, as well as scalable enough to support large number of identity document types. The aim of the presented framework is to serve as a basis for developing new methods and algorithms for ID document recognition, as well as for far more heavy challenges of identity document forensics, fully automated personal authentication and fraud prevention.

A review on existing learning techniques applied in solving optical character recognition problem

Optical character recognition is treated as the classical problem in image processing that can be widely used for application areas related to computer vision. This process is highly dependent on the languages, as well as various formats of information like hand written/printed/sculptured documents as input. The chief operations that can be performed while handling this task are training the system and predicting the characters based on the trained knowledge. Conventionally, classification and regression mechanisms are used for this purpose. Due to the recent advents in technological progress, the same mechanism was served with the help of modern machine learning and deep learning concepts. The main agenda of this paper is to create awareness among the researchers by discussing existing mechanisms related to solve optical character recognition using various machine learning and deep learning concepts.

A Review on Existing Learning Techniques Applied in Solving Optical Character Recognition Problem

Optical character recognition is treated as the classical problem in image processing that can be widely used for application areas related to computer vision. This process is highly dependent on the languages, as well as various formats of information like hand written/printed/sculptured documents as input. The chief operations that can be performed while handling this task are training the system and predicting the characters based on the trained knowledge. Conventionally, classification and regression mechanisms are used for this purpose. Due to the recent advents in technological progress, the same mechanism was served with the help of modern machine learning and deep learning concepts. The main agenda of this paper is to create awareness among the researchers by discussing existing mechanisms related to solve optical character recognition using various machine learning and deep learning concepts.

Implementation of Optical Character Recognition using Tesseract with the Javanese Script Target in Android Application

Recognising characters from text have been a popular topic in the computer vision area. The application can benefit to many problems in the world. For example: recognising text in documents, classifying the text or scripts of documents, plate recognition, etc. Many researchers have been developed the methods for recognising characters in by using Optical Character Recognition methods. Although text recognition problem using Optical Character Recognition has been more or less solved, most of the Optical Character Recognition problem explored is belong to Latin alphabet texts. Meanwhile, there are several languages have non-Latin scripts as the written text. Recognising a non-Latin script is quite challenging as the contour and shape of the text are relatively different with a Latin script text. This research aims to collect datasets for OCR in Javanese characters. A total of 5880 characters were collected and trained with several methods with Tesseract OCR tools. The models then be implemented to a mobile phone (Android based). The highest accuracy (97,50%) achieved by the model was achieved by combining single boundary box for the whole parts of the character and the separate boundary boxes in main body and sandangan parts.

Mass Digitization of Early Modern Texts With Optical Character Recognition

Optical character recognition (OCR) engines work poorly on texts published with premodern printing technologies. Engaging the key technological contributors from the IMPACT project, an earlier project attempting to solve the OCR problem for early modern and modern texts, the Early Modern OCR Project (eMOP) of Texas A8M received funding from the Andrew W. Mellon Foundation to improve OCR outputs for early modern texts from the Eighteenth Century Collections Online (ECCO) and Early English Books Online (EEBO) proprietary database products—or some 45 million pages. Added to print problems are the poor quality of the page images in these collections, which would be too time consuming and expensive to reimage. This article describes eMOP's attempts to OCR 307,000 documents digitized from microfilm to make our cultural heritage available for current and future researchers. We describe the reasoning behind our choices as we undertook the project based on other relevant studies; discoveries we made; the data and the system we developed for processing it; the software, algorithms, training procedures, and tools that we developed; and future directions that should be taken for further work in developing OCR engines for cultural heritage materials.

Machine learning and empathy: the Civil Rights CAPTCHA

Human interactive proofs HIPs are a basic security measure on the Internet to avoid automatic attacks. There is an ongoing effort to find a HIP that is secure enough yet easy for humans. Recently, a new HIP has been designed aiming at higher security: the Civil Rights Completely Automated Public Turing test to tell Computers and Humans Apart CAPTCHA. It employs the empathy capacity of humans to further strengthen Securimage, a well-known text CAPTCHA. In this paper, we analyze it from a security perspective, finding fundamental design flaws. Using several well-known machine learning ML algorithms, we analyze to what extent these flaws affect its security. We discover that thanks to them, we can create a successful side-channel attack. This attack is able to correctly solve the HIP on 20.7% of occasions, much more than enough to consider it broken. Thus, we show that there is no need to solve the problem of optical character recognition nor empathy analysis for computers to break this particular HIP. ML can be successfully used to break a HIP that uses both with a side-channel attack. This security analysis can be applied to other HIPs. It will allow to test whether they are leaking too much information by unexpected ways, given non-evident design flaws. Copyright © 2015 John Wiley & Sons, Ltd.

Arabic character recognition using a Haar cascade classifier approach (HCC)

Optical character recognition (OCR) shows great potential for rapid data entry, but has limited success when applied to the Arabic language. Traditional OCR problems are compounded by the nature of Arabic language and because the script is heavily connected. A machine learning, Haar cascade classifier (HCC) approach was introduced by Viola and Jones (Rapid object detection using a boosted cascade of simple features. Kauai, Hawaii, 2001) to achieve rapid object detection based on a boosted cascade of simple Haar-like features. Here, that approach is applied for the first time to suit Arabic glyph recognition. HCC approach eliminates problematic steps in the pre-processing and recognition phases and, most importantly, character segmentation stage. A classifier was produced for each of the 61 Arabic glyphs that exist after the removal of diacritical marks (dots). These classifiers were trained and tested on some 2,000 images each. The system was tested with real text images and produces a recognition rate for Arabic glyphs of 87 %. The technique gives good results relative to those achieved using a commercial Arabic OCR application and existing state-of-the-art research application.

Support vector machine for identification of handwritten Gujarati alphabets using hybrid feature space

Gujarati language is used in the western state of Gujarat. Because of its peculiarities, its Optical Character Recognition becomes very difficult. For this language very less work has been done in the area of Optical Character Recognition. In this paper, I have attempted the problem of Optical Character Recognition for handwritten Gujarati alphabets. For this work, forty handwritten alphabets are collected from about one hundred and ninety nine writers. Here; aspect ratio, extent of alphabet, and image subdivision approach has been used as feature space and support vector machine (SVM) has been used for the classification purpose and it gives 86.66 % of performance accuracy. kNN is also used for classification and the result is compared with that of SVM. The paper also describes support vector machine.

Loose, Falling Characters and Sentences: The Persistence of the OCR Problem in Digital Repository E-Books

The electronic conversion of scanned image files to readable text using optical character recognition (OCR) software and the subsequent migration of raw OCR text to e-book text file formats are key remediation or media conversion technologies used in digital repository e-book production. Despite real progress, the OCR problem of reliability and accuracy in OCR-derived e-book text and metadata persists. This paper examines a selection of digitized e-books in several prominent digital repositories and discusses the impact of OCR technology on e-book text file formats, metadata, and the online reading experience.

Precise-Spike-Driven Synaptic Plasticity: Learning Hetero-Association of Spatiotemporal Spike Patterns

A new learning rule (Precise-Spike-Driven (PSD) Synaptic Plasticity) is proposed for processing and memorizing spatiotemporal patterns. PSD is a supervised learning rule that is analytically derived from the traditional Widrow-Hoff rule and can be used to train neurons to associate an input spatiotemporal spike pattern with a desired spike train. Synaptic adaptation is driven by the error between the desired and the actual output spikes, with positive errors causing long-term potentiation and negative errors causing long-term depression. The amount of modification is proportional to an eligibility trace that is triggered by afferent spikes. The PSD rule is both computationally efficient and biologically plausible. The properties of this learning rule are investigated extensively through experimental simulations, including its learning performance, its generality to different neuron models, its robustness against noisy conditions, its memory capacity, and the effects of its learning parameters. Experimental results show that the PSD rule is capable of spatiotemporal pattern classification, and can even outperform a well studied benchmark algorithm with the proposed relative confidence criterion. The PSD rule is further validated on a practical example of an optical character recognition problem. The results again show that it can achieve a good recognition performance with a proper encoding. Finally, a detailed discussion is provided about the PSD rule and several related algorithms including tempotron, SPAN, Chronotron and ReSuMe.

Probabilistic management of OCR data using an RDBMS

The digitization of scanned forms and documents is changing the data sources that enterprises manage. To integrate these new data sources with enterprise data, the current state-of-the-art approach is to convert the images to ASCII text using optical character recognition (OCR) software and then to store the resulting ASCII text in a relational database. The OCR problem is challenging, and so the output of OCR often contains errors. In turn, queries on the output of OCR may fail to retrieve relevant answers. State-of-the-art OCR programs, e.g., the OCR powering Google Books, use a probabilistic model that captures many alternatives during the OCR process. Only when the results of OCR are stored in the database, do these approaches discard the uncertainty. In this work, we propose to retain the probabilistic models produced by OCR process in a relational database management system. A key technical challenge is that the probabilistic data produced by OCR software is very large (a single book blows up to 2GB from 400kB as ASCII). As a result, a baseline solution that integrates these models with an RDBMS is over 1000x slower versus standard text processing for single table select-project queries. However, many applications may have quality-performance needs that are in between these two extremes of ASCII and the complete model output by the OCR software. Thus, we propose a novel approximation scheme called Staccato that allows a user to trade recall for query performance. Additionally, we provide a formal analysis of our scheme's properties, and describe how we integrate our scheme with standard-RDBMS text indexing.

Optical character recognition: A comprehensive study of hybrid methods

The problem of Optical Character Recognition (OCR) is methodically treated in this paper. The paper discusses some traditional methods like template matching and neural networks that are employed by the more advanced and artificially intelligent systems. Though K-PCA, a non-intelligent system, prov ides better accuracy, it was found to occur at the cost of more processing time. Neural Networks provide a trade-off between processing time and accuracy.

A hierarchical approach to recognition of handwritten Bangla characters

A novel hierarchical approach is presented here for optical character recognition (OCR) of handwritten Bangla words. Instead of dealing with isolated characters as found in selected works [T.K. Bhowmik, U. Bhattacharya, S.K. Parui, Recognition of Bangla handwritten characters using an MLP classifier based on stroke features, in: Proceedings of the ICONIP, Kolkata, India, 2004, pp. 814–819; K. Roy, U. Pal, F. Kimura, Bangla handwritten character recognition, in: Proceedings of the Second Indian International Conference on Artificial Intelligence (IICAI), 2005, pp. 431–443; S. Basu, N. Das, R. Sarkar, M. Kundu, M. Nasipuri, D.K. Basu, Handwritten Bangla alphabet recognition using an MLP based classifier, in: Proceedings of the Second National Conference on Computer Processing of Bangla, Dhaka, 2005, pp. 285–291; A.F.R. Rahman, R. Rahman, M.C. Fairhurst, Recognition of handwritten Bengali characters: a novel multistage approach, Pattern Recognition 35, 2002, pp. 997–1006; U. Bhattacharya, S.K. Parui, M. Sridhar, F. Kimura, Two-stage recognition of handwritten Bangla alphanumeric characters using neural classifiers, in: Proceedings of the Second Indian International Conference on Artificial Intelligence (IICAI), 2005, pp. 1357–1376; U. Bhattacharya, M. Sridhar, S.K. Parui, On recognition of handwritten Bangla characters, in: Proceedings of the ICVGIP-06, Lecture Notes in Computer Science, vol. 4338, 2006, pp. 817–828], the present approach segments a word image on Matra hierarchy, then recognizes the individual word segments and finally identifies the constituent characters of the word image through intelligent combination of recognition decisions of the associated word segments. Due to possible appearances of consecutive characters of Bangla words on overlapping character positions, segmentation of Bangla word images is not easy. For successful OCR of handwritten Bangla text, not only recognition but also segmentation of word images are important. In this respect the present hierarchical approach deals with both segmentation and recognition of handwritten Bangla word images for a complete solution to handwritten word recognition problem, an essential area of OCR of handwritten Bangla text. In dealing with certain category of word segments, created on Matra hierarchy, a sophisticated recognition technique, viz., two-pass approach [S. Basu, C. Chaudhury, M. Kundu, M. Nasipuri, D.K. Basu, A two pass approach to pattern classification, in: N.R. Pal et al. (Ed.), Lecture Notes in Computer Science, vol. 3316, ICONIP, Kolkata, 2004, pp. 781–786] is employed here. The degree of sophistication of the classification technique is also rationally tuned depending on various categories of word segments to be recognized. For example, the two-pass approach is employed here for recognizing middle zone character segments, whereas recognition of middle zone modified shapes of Bangla script is done through simple template matching. Considering learning and generalization abilities of multi layer perceptrons (MLPs), MLP based pattern classifiers are used here for most of the classification related tasks. A powerful feature set is also designed under this work for recognition of complex character patterns using three types of topological features, viz., longest-run features, modified shadow features and octant-centroid features. In a nutshell, the work deals with a practical problem of OCR of Bangla text involving recognition as well as segmentation of constituent characters of handwritten Bangla words.

Training RBF networks with selective backpropagation

Backpropagation with selective training (BST) is applied on training radial basis function (RBF) networks. It improves the performance of the RBF network substantially, in terms of convergence speed and recognition error. Three drawbacks of the basic backpropagation algorithm, i.e. overtraining, slow convergence at the end of training, and inability to learn the last few percent of patterns are solved. In addition, it has the advantages of shortening training time (up to 3 times) and de-emphasizing overtrained patterns. The simulation results obtained on 16 datasets of the Farsi optical character recognition problem prove the advantages of the BST algorithm. Three activity functions for output cells are examined, and the sigmoid activity function is preferred over others, since it results in less sensitivity to learning parameters, faster convergence and lower recognition error.

Premature clustering phenomenon and new training algorithms for LVQ

Five existing LVQ algorithms are reviewed. The Premature Clustering Phenomenon, which downgrades the performance of LVQ is explained. By introducing and applying the “ equalizing factor” as a remedy for the premature clustering phenomenon a breakthrough is achieved in improving the performance of the LVQ network, and its performance becomes competitive with that of the best known classifiers. For estimating the equalizing factor four different formulas are suggested, which result in four different versions of the LVQ4a algorithm. A new weight-updating formula for LVQ is presented, and the LVQ4b algorithm is presented as implementation of this new weight-updating formula in batch mode training. In addition, four variants of the LVQ4c algorithm are presented as the customized LVQ4b algorithm for pattern mode training. A meticulous analysis of their performances and that of five early training algorithms has been carried out and they have been compared against each other, on 16 databases of the Farsi optical character recognition problem.

A segmentation-free approach to text recognition with application to Arabic text

As the recognition rates and speeds of optical character recognition (OCR) systems steadily improve, the problem of OCR--and subsequently research interest--is shifting from recognizing: isolated, high-quality characters to reading cursive scripts and degraded documents. In recognizing such texts, a major undertaking is segmenting cursive words into characters and isolating merged characters. In OCR systems that recognize cursive text, the segmentation subsystem becomes the pivotal stage in the system to which a sizable portion of processing is devoted and a considerable share of recognition errors is attributed. The most notable feature of Arabic writing is its cursiveness. It also poses the most difficult problem for recognition algorithms. In this work, we describe the design and implementation of a system that is automatically trainable and that recognizes noisy and cursive words. To recognize a word, the system does not segment it into symbols (character shapes) in advance; rather, it recognizes the input word by detecting a set of shape primitives on the word. It then matches the regions of the word (represented by the detected primitives) to a set of symbol models. A spatial arrangement of symbol models that are matched to regions of the word, then, becomes the description of the recognized word. Since the number of potential arrangements of all symbol models is large, the system imposes a set of word structure and spatial consistency. It searches the space comprised of the arrangements that satisfy the constraints and tries to maximize the a posteriori probability of the symbol-models' arrangement. Large-scale experimentation with the system on isolated characters reveals that it has a recognition rate of 99.7% for synthetically degraded symbols and 94.1% for scanned symbols. Experimentation on isolated words reveals that the system has a recognition rate of 99.4% for noise-free words, 95.6% for synthetically degraded words, and 73% for scanned words. The main theoretical contribution of this work is in laying the foundation for a segmentation-free approach for Arabic word recognition. Recognition is based on maximizing the probability of the word given the detected primitives. The system is designed to minimize training effort and is extensible as training determines the symbols the system recognizes.

Multiple subclass pattern recognition: A maximin correlation approach

This paper addresses a correlation based nearest neighbor pattern recognition problem where each class is given as a collection of subclass templates. The recognition is performed in two stages. In the first stage the class is determined. Templates for this stage are created using the subclass templates. Assignment into subclasses occurs in the second stage. This two stage approach may be used to accelerate template matching. In particular, the second stage may be omitted when only the class needs to be determined. The authors present a method for optimal aggregation of subclass templates into class templates. For each class, the new template is optimal in that it maximizes the worst case (i.e. minimum) correlation with its subclass templates. An algorithm which solves this maximin optimization problem is presented and its correctness is proved. In addition, test results are provided, indicating that the algorithm's execution time is polynomial in the number of subclass templates. The authors show tight bounds on the maximin correlation. The bounds are functions only of the number of original subclass templates and the minimum element in their correlation matrix. The algorithm is demonstrated on a multifont optical character recognition problem. >